correlation between FEM and measured data

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

goslings - 24 Jun 2007 12:15 GMT

I have posted a longer version in MATLAB newsgroup
Anyone know a good correlation technique for FEM displacement v
Measured displacements
that produces an overall single number (0 - 1) ?

Reply to this Message

Ben - 24 Jun 2007 16:09 GMT

> I have posted a longer version in MATLAB newsgroup
> Anyone know a good correlation technique for FEM displacement v
> Measured displacements
> that produces an overall single number (0 - 1) ?

Interesting problem.

I spent some time on Google and came up with a good reference and with
some Matlab code that may suit your needs.

http://ncmi.bcm.tmc.edu/ncmi/publication/publications/papers/2003Cross-correlati
on_and_merging_of_crystallographic_reflections_derived_from_cryoelectron_microgr
aphs_of_3D_crystals--applications_to_the_limulus_acromosomal_bundle.pdf

The paper above doesn't go into the details of how they calculate
their 3D cross-correlation, but they do give references.

Here's some Matlab code developed for medical imaging.

http://www.cs.ubc.ca/~deaton/tut/normxcorr3.html

Reply to this Message

goslings - 24 Jun 2007 18:17 GMT

> > I have posted a longer version in MATLAB newsgroup
> > Anyone know a good correlation technique for FEM displacement v
[quoted text clipped - 14 lines]
>
> http://www.cs.ubc.ca/~deaton/tut/normxcorr3.html

cheers
I'll check them out
Ian

Reply to this Message

Math Forum / Mathematics / Numerical Analysis / June 2007

correlation between FEM and measured data